Anthropic releases cheaper AI model ahead of its blockbuster public listing

By the Editorial Team · Wednesday, September 23, 2026

Anthropic launched Claude Opus 5.5, an AI model offering similar performance to its previous flagship but at about 40% lower cost, targeting business customers. This pricing strategy aims to boost enterprise adoption ahead of the company's major IPO this autumn as it faces increasing competition from rivals like OpenAI and emerging open-weight AI models.

Anthropic unveiled a new artificial intelligence model called Claude Opus 5.5 on Wednesday, promoting it as delivering comparable performance to the company’s most advanced tool while significantly reducing costs. This move comes as the San Francisco-based AI firm prepares for a highly anticipated initial public offering (IPO) slated for this autumn, which is expected to be one of the largest ever.

The newly released model is approximately 40% more affordable than its predecessor, positioning it as the first in a forthcoming “family” of cost-efficient models aimed at businesses, Anthropic said. Claude Opus 5.5 will be priced at $4 per million input tokens and $20 per million output tokens, compared to the company’s prior model, Fable 5.1, which costs $10 per million input tokens and $50 per million output tokens. Tokens are the units of data that constitute AI queries.

Anthropic is seeking to expand its enterprise customer base and demonstrate sustained growth amid intensifying competition and a market increasingly sensitive to pricing. Investors have speculated that the launch of this new model might trigger the release of the company’s IPO prospectus. The firm is navigating a competitive landscape that includes OpenAI, which recently reduced its own AI pricing to stimulate growth after a slower start this year. OpenAI also promotes its latest model, Astra, as the most powerful on the market.

Both Anthropic and OpenAI face growing challenges from so-called “open-weight” models—AI systems that users can customize and operate on personal hardware. Many of these models originate from China and have reportedly captured market share from U.S. AI developers, according to data from cloud services provider Vercel.

The release of Opus 5.5 follows remarks from Anthropic’s CEO Dario Amodei, who called on AI labs to moderate the pace of development in an effort to reduce risks related to security breaches and the broader existential threats posed by AI technologies. The company stated that the new model underwent extensive alignment testing and pre-release evaluation by independent organizations, incorporating safeguards aimed at sensitive domains such as cybersecurity and biology.

As Anthropic approaches its IPO, market observers will be watching closely to see whether the company can maintain its rapid growth trajectory while contending with pricing pressures and intensifying competition in the global AI sector.

This article was reported by synthesizing coverage from multiple independent news outlets to present a balanced, unbiased account. Read about our editorial standards. Spotted an error? Let us know.